Do I need to scale test data and Dependent variable in the train data?

I am new to the concept of scaling a feature in Machine Learning, I read that scaling will be useful when one feature range is very high when compared to other features. But if I choose to scale the training data then:

1. Can I just scale that one feature that has high range?
2. If I scale the entire X of train data then do I need to also scale the y of train data and entire test data?